Deliverable D19: Influence of alluvium filled basins and edge effects on displacement response spectra

Abstract

Several cases of strong motion data recorded in alluvial basins show very high amplification effects that exceed, at medium and long periods, those predicted by empirical relations or included in standard reference codes. Among others, we recall here the earthquakes of San Fernando (1972), Northridge (1994) (BSSA, 1996) and the Umbria-Marche sequence of 1997, during which the shaking recorded in the Gubbio Plain (Pacor et al, 2007) provided strong evidence of the phenomenon: extremely high spectral amplification values (higher than reference codes) for 2 < T < 4 s (see Fig. 4). The main reason why these amplifications (“basin effects”) take place lies in the generation, by diffraction at the edges of the valley, of long period surface waves that travel horizontally in the upper sediments of the valley. While the influence of alluvium filled basins on site response has been the subject of substantial research, the resulting modifications on the response spectra at the basin surface (especially for T > 1 - 2 s) has not been as thoroughly investigated (see e. g. Chávez e Faccioli, 2000) despite its importance in structural design. Significant previous studies tried to estimate basin amplification effects through the analysis of strong-motion data and most of them quantified basin geometry only in terms of sediment depth (Trifunac and Lee, 1978), introducing such term in newly developed attenuation models (Campbell, 1997; Field, 2000; Lee and Anderson, 2000). Other studies tried to relate basin effects also to the relative location of source and site position in the basin (Choi et al., 2005), or to the distance to the basin edge (Joyner, 2000). In the S5 project, specific parametric studies involved two different types of basins (both typical of the Italian Peninsula): “enclosed” basins and “open” basins. Results gave great insight of how basin effects amplify seismic motion in connection with the geometry of the basin, with the fault mechanism and with the different valley-fault configurations. Critical in all analyses is the value of the fundamental 1D vibration period of valley centre, T01D, that acts as the theoretical upper limit to 2D basin amplification effects, as stated in Chàvez-García and Faccioli (2000). Its calculation appears thus crucial in the study of the seismic response of valleys and basins. In addition, a detailed study has been devoted to the identification and classification of alluvium filled basins in Italy, and particularly to the parameters that the previous analyses highlighted as the most critical ones in the modification of response spectra.